Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=Transcript of the talk, from the video's captions. Auto-generated: speaker names in particular are unreliable.
=

# Menstrual Stigma and Human Capital: Experimental Evidence from Madagascar

Authors:   
Discussant: None  
Video: https://www.youtube.com/watch?v=PHYYOyrSItw&t=18167s


## Talk (05:02:47 – 05:12:50)

[05:02:47] >> Okay. So uh we go straight to the lightning round. So we're going to have three uh talks >> of nine minutes each. Uh no questions.  
[05:02:57] The first one uh is by Duncan Webb. >> Hi everyone. Thanks so much. It's nice to be back at the MBR um summer institute. So my name is Duncan Webb and this is joint work with Karen McCor and  
[05:03:11] Julieta and this is a project that tries to understand how does menstruation potentially act as a barrier to human capital development for adolescent girls in a low-inccome context in this uh in this setting uh Madagascar.  
[05:03:24] So menstruation is a very widespread phenomenon. I don't need to tell anyone that. Um and lots of people who menstruate are can risk shame and harassment because of the taboo and the stigma that surrounds it as a topic. And we think that menstruation might act as  
[05:03:39] a barrier to human capital development for for teenage girls um through a variety of channels. One cluster of channels I'm going to call conventional channels, which is stuff people have thought about in the context of menstruation before. Maybe it's the case that girls in low-income settings have  
[05:03:53] poor access to menstrual products or don't know how to enact good hygiene practices and therefore their health is worse or potentially they are reluctant or not able to go to school to attend school when they're menstruating. That's things people have thought about a bit  
[05:04:08] more before that we think there's also a set of what we're going to call psychosocial channels related to menstruation um because of the stigma and the taboo that surrounds it as a topic that could lead to human capital development. uh being more difficult for  
[05:04:21] girls as well. If you think about girls being stressed, being more likely to be teased, a worse school environment at school, those could also inhibit learning even when girls are are attending school the same amount.  
[05:04:34] And we think that the research on this topic uh leaves quite a lot of uncertainty about the effects of menstruation related uh interventions.  
[05:04:41] So typically interventions have focused on distributing free uh sanitary products. Um but very little research has looked at uh the effect not only on attendance but also on directly on learning and very little research has  
[05:04:55] focused on the stigma and the taboo that surrounds menstruation and how that could affect on its own uh girls schooling.  
[05:05:03] So our our research questions in this context are going to be can addressing menstruation related barriers improve learning for girls in a low-inccome context and if so through what mechanisms and so what we do in this paper is run  
[05:05:17] an RCT in 140 schools in rural Madagascar in which we're evaluating the effect of a multifaceted hygiene and menstruation related intervention. That intervention is going to include both what we're going to call hard  
[05:05:30] components. Um so giving people access to the materials they need to manage their menstruation. So by building latrines and wash basins and also distributing sanitary pads but also soft components by which I mean sensitization  
[05:05:43] on hygiene and menstruation and trying to reduce the stigma that surrounds menstruation as well. And there throughout this whole program there's a very strong focus on the stigma aspect in particular. And so to preview where I'm going to go with the results, we  
[05:05:58] find that this program leads to pretty substantial improvements in learning uh for adolescent girls in these schools.  
[05:06:04] And interestingly enough, against our own prior and many other people's prior, that does not seem to be explained by girls going to school more, increased attendance or better health. And instead, we find that the channels that  
[05:06:16] do move are kind of psychosocial. So we find these very strong reductions in stigma uh menstrual stigma. And we find pos possibly related reductions in stress for girls and improvements in the  
[05:06:30] school climate as well. So we're doing a rand uh randomization at the school level where 35 schools uh don't receive an intervention. Um some schools receive what we're calling the base program which is a combination of  
[05:06:43] as I said the hard and soft components. So the hard the soft components include teacher sensitization where um a subset of the teachers in each school are trained on a curriculum of hygiene and menstrual hygiene and reducing uh menstrual stigma and then they're asked  
[05:06:58] to sensitize the the students in their class on a regular basis. There's also a kind of school level mobilization involving teachers, students and um and parents as well uh through a kind of hygiene committee and a school level  
[05:07:11] contest. And then there's the hard components as well. So we build uh latrines and wash basins in the schools in the treatment schools and distribute free reusable sanitary pads to girls in the schools as well.  
[05:07:25] In some schools we also layer on top of that in uh base uh intervention uh the young girl leaders intervention which is basically selecting around four girls per school to act as ambassadors for menstrual hygiene and reducing menstrual  
[05:07:38] stigma. They're selected to be what is in the literature called positive deviants. people willing to openly speak out against the norm of taboo around menstruation. Um, and then they're uh trained and uh coached to sensitize  
[05:07:51] their peers on this topic and to reduce the stigma on the topic as well.  
[05:07:56] So what do we find? So I think there's no questions because it's very sorry. Uh so what do we find? We find that as I said before this program the base only program leads to pretty substantial improvements in learning for girls in  
[05:08:11] these schools. So we we we carry out standardized learning tests combining math, literacy and cognitive skills and find a roughly 02 standard deviation improvement for girls in the treatment schools. That also translates to um  
[05:08:25] sorry that that that intervention actually sits that that effect size sorry sits at the 75th percentile of learning programs evaluated in low and middle inome countries. So, we think that's pretty substantial and it also  
[05:08:38] translates to uh increased progression through grades from about 52% of girls to 63% of girls between the year 1 and year two of the program. It also translates to improvements in girls official GPA uh from the the school  
[05:08:52] admin data. And interestingly enough, we don't find an equivalent improvement for boys GPA although that difference is not significantly uh uh detectable across gender.  
[05:09:04] We also find no additional improvements in learning driven by the young girl leaders program. There's plenty more in the paper on why we think that is. One possible explanation is that in primary schools in particular, it's very difficult to identify girls who are  
[05:09:16] willing to openly go against an existing uh uh taboo and there some evidence of of crowd out as well.  
[05:09:24] So the program itself does not focus on academic performance. It does not focus on math or literacy and yet we find these learning gains. So why do we think learning improved in the treatment school? So the obvious candidate and as I said what previous researchers focused  
[05:09:39] on is that girls are maybe healthier because they're enacting good hygiene practices um and they are have access to toilets and sanitary products. But we don't find any evidence of this kind of uh channel or any evidence of increased  
[05:09:52] attendance as well. So we find no measurable impacts on enrollment or attendance or phys physical health. And there's some other suggestive evidence as well that these types of channels are not at play. For example, the grade  
[05:10:05] progression. So moving from grade one uh between years one and two progressing through the grade that happens before the infrastructure was even complete. Um and we also find that the learning even improves for girls that have not started  
[05:10:18] men uh uh yet in the in the sample suggesting that it's other types of channels that are at play.  
[05:10:25] So what do we find evidence of instead is a possibly interrelated set of psychosocial channels um that could be explaining the the learning improvements. The first is we find these very strong impacts on menstrual stigma which we measure through a set of survey  
[05:10:40] measures and also a battery of lab in the field measures which we developed to try and understand the revealed preference behavior related to to menstrual stigma. We also and possibly relatedly we then uh show that girls appear to be less stressed in in  
[05:10:55] treatment schools. Um so we measure their heart rate at endline and find a reduction in the average heart rate um which we interpret as a a biomarker of reduced stress and then we find evidence of an overall improvement in the school  
[05:11:10] environment. The psychosocial climate at school where severe bullying is going down in treatment schools and peer connections. So girls uh being friends and collaborating with other other students uh improves. There's one important adverse effect that is  
[05:11:24] possibly related to that improvement in the school environment which is that the base only program actually raises pregnancy rates in the sample from two to three and a half%. Um but there's no equivalent increase in the when the younger leaders are present. So that seems to be an important benefit of the  
[05:11:39] younger leaders. Um and we think that's probably related to the fact that girls and boys seem to be more friends with each other. and you can see what happens next.  
[05:11:47] Um, we think these together could be um it could explain the um the contribute to the improvements in learning through for example better concentration at school if you're less stressed and less bullied, improved overall quality of  
[05:12:01] time at school and uh better motivation as well um in school work. So to conclude, we find this intervention focused on menstruation and menstrual hygiene leads to these large learning improvements in schools. We don't find  
[05:12:15] evidence that that's driven by improved health or improved attendance, but instead by these psychosocial channels, so lower stress, better school climate, and reduce stigma around menstruation.  
[05:12:25] And so we're in the process of trying to work with the ministry in the World Bank to uh scale this up uh to the majority of secondary schools in in Madagascar.  
[05:12:34] And more broadly, we think that this indicates that menstruation and menstrual stigma could be an important barrier to girls schooling in even in a low-inccome context like Madagascar.  
[05:12:44] Thanks very much and looking forward to confess.